Atypical hyperplasia of the breast (atypia) is a common condition among women, showing an increasing trend in terms of number of cases per year. It is well known that women presenting with atypia are at markedly increased risk of subsequently developing invasive breast cancer. However, risk varies widely among these women, depending on environmental and behavioral risk factors, family history, use of chemopreventative agents or preventative surgery. Providing accurate estimates of absolute risk of invasive breast cancer for women with atypia remains a challenge. Existing risk prediction tools address some aspects of benign breast disease and have been a valuable starting point, but they remain somewhat insufficient: they only include an overall adjustment for a prior diagnosis of benign breast disease, are not trained on large and well characterized atypia cohorts, and do not address the causal effect of chemotherapy in a prospective fashion. To overcome this problem, we have built an absolute risk prediction model to assist the clinical management of women diagnosed with atypia. Aims. The goals of this application are to refine this model via the inclusion of detailed pedigree information, validate it in two independent cohorts, and disseminate through widely used clinical decision support tools. These goals will be achieved through the following specific aims: 1. Extension and Validation of a model for predicting invasive breast cancer risk among women with atypical hyperplasia of the breast; specifically: 1a. Model extension, to incorporate consideration of a detailed description of family history of breast cancer and to incorporate the effect of mastectomy; 1b. Model Validation, to validate the resulting model in two independent cohorts of women with atypia. 2. Implementation of software tools for the clinical application of the model from Aim 1. These tools will be disseminated in tree complementary ways: a) WebRiskService, a free access web service that currently serves as the computational engine for HughesRiskApps and other primary care decision support tools; b) CancerGene, a freely distributed genetic counseling tool used by several thousand clinics worldwide; c) BayesMendel, a standalone open source statistical software for mendelian risk prediction in the R language. In addition, through the BayesMendel package, the tools developed in this aim will be used to enhance the capabilities of the existing carrier probability model BRCAPRO. Impact. These aims will provide both a rigorous epidemiological foundation and practical tools for the clinical management of women with atypia. Use of these tools will increase women's awareness of their risk and allow a more rational and objective evaluation of available treatment options, through a more careful consideration of the balance between the possible benefits and the possible side effects of prevention approaches.